Twenty years of gas chromatographs at QLD underground coal mines – Part two

Source: Queensland Mines and Energy

The initials MTI were derived from an earlier version of the company name, Microsensor Technology Incorporated. The micro GC, initially developed for NASA to measure the constituents of the Martian atmosphere, featured much faster analysis times. Typically, these were less than two minutes. The first of these systems was installed at Moranbah North in 1997.
The micro GC utilised in this system incorporated four individual modularised GCs.  This allowed separate GC modules to be dedicated to the analysis of particular gases, with no need for any one of them to measure all of the gases.
Similar to previous generations of Camgas, molecular sieve and porous polymer columns were also utilised in this Camgas configuration. The biggest difference was the introduction of capillary columns with an internal diameter of 0.32 mm. The advantages of capillary columns over standard packed columns are well known amongst chromatographers. Because of the four module configuration no column switching was required, simplifying the operation of the system. Added to these advantages, there was no longer a requirement for the columns to be matched.
The micro GC utilised TCDs capable of much better detection limits than previous detectors of this type. The detection of hydrogen was so much improved in these GCs that it become evident that it was normal to see hydrogen in most samples collected from an underground coal mine. Previously, because of much higher detection limits, it was believed that the presence of hydrogen indicated a spontaneous combustion event was occurring as it was only during such times, sufficient concentrations of hydrogen were generated to exceed the detection thresholds of the older generation instruments. This resulted in a need to re-educate the whole industry that some hydrogen was normal and that a baseline concentration needed to be established at each mine to watch for deviations from the baseline.
Because of the multiple GC modules, carrier gases could be selected for individual modules dependent on the sample gases to be measured. This further enhanced detection limits. The lack of an FID in the system eliminated the requirement for the fuel gas and air supply and the problems associated with setting and maintaining the correct ratios of these. The requirement for a methaniser was also eliminated. The overall outcome of a system without an FID and a much more sensitive TCD was a significantly less complex GC to operate with reduced maintenance requirements. All of this came without significant compromise to sensitivity.
Another advantage that came with the micro GCs that cannot be overstated was the ability to easily transport them to a mine site and ready them for operation in minimal time to provide complete gas analysis during a mine emergency. This provision was just not possible with the previous conventional GCs.
Although the micro GC was initially manufactured by MTI Analytical Instruments, in 1998 they were bought out by Hewlett Packard. Hewlett Packard was established in 1939 by Bill Hewlett and Dave Packard as a company manufacturing test and measurement equipment. It was not until 1966 that Hewlett Packard ventured into the computing world with which their name has become so synonymous. In 1999 Hewlett Packard split its computing and test and measurement divisions into two separate companies, with the test and measurement company becoming Agilent Technologies.
Agilent Technologies revised the MTI micro GC design in 2004 when it released the new Agilent 3000 micro GC. There were a range of improvements including field swappable plug and play modules and Ethernet communications. This became the fourth generation Camgas GC and was first installed at North Goonyella in 2004.
The current Camgas system features the latest 3000 micro GC offered by Agilent which incorporates an improved detector and module layout and configuration and electronic pressure controls, but to the end user operates in a similar fashion to the original micro GC. The virtual chemist has moved into the 21st century.

Sample treatment is critically important
The various techniques by which samples are collected and contained, coupled with the use of pressurised calibration gases, has created challenges in presenting representative and repeatable samples to the GC. To overcome this problem, Simtars developed a separate sample introduction system capable of conditioning the sample to remove contaminants capable of compromising the operation of the GC as well as handling samples at atmospheric and elevated pressures.
One of the most significant challenges faced in designing a suitable mine gas analysis system based on chromatography has been the need to deal with the wide range of gas concentrations and sample compositions coming from the underground coal mine. Samples can vary from fresh air all the way through to 100 per cent methane containing varying amounts of other typical underground mine gases in ppm to percent ranges. Chromatograph method configurations and specified calibration gases must be capable of covering the full range of gases and gas concentrations likely to be encountered.

Not just for use in an emergency
Although the Moura No. 4 Inquiry indicated that the GCs were required for gas analysis following a mine disaster some mines adopted a more proactive approach. They started using them as a tool to ensure that no indication of spontaneous combustion was evident and for analysis of samples that were not covered by the mine’s automated monitoring system.
The Inquiry into the Moura No 2 explosion that occurred on August 7th, 1994, recommended that Simtars promote the use of GCs for routine (non emergency) analysis of gas samples.
Although the information gained from GC analysis could be invaluable in the interpretation of the underground environment, the long analysis times (20-30mins) of the first systems did not lend themselves to the analysis of large numbers of samples. Furthermore, the sensitivity of these earlier instruments was not necessarily good enough to detect the early onset of spontaneous combustion. Hence the perception by some mines at the time was that GC was a technique with the primary purpose of being used in a mine emergency.
With the improvements in technology already outlined, analysis time has dropped to several minutes and detection limits have improved greatly. This along with Simtars encouraging a proactive approach to gas analysis by GC has resulted in mines conducting routine assessments of the underground environment using the technique. As such the role of the GC has shifted from gas analysis during emergencies to routine assessment of the underground environment.
Queensland mines recognise GC analysis as the only technique that provides complete analysis of gas composition.  Because of this, samples are regularly analysed by GC to provide mine personnel with the best possible information to make informed assessments of the underground environment. Samples are routinely analysed to look for the presence of hydrogen and ethylene, both indicators of spontaneous combustion and both only measureable by GC.
Gas chromatography also provides a tool for analysis of samples not covered by the mine’s automated monitoring system. Samples can be collected from throughout the mine and analysed using the GC. It is difficult to guess how many significant events would have eventuated if GCs had not been used for routine analysis as many “low level” incidents have been identified early enough for controls to be implemented successfully, often with minimal disruption to the mine’s operation.
Ironically the GC has become used so much for routine analysis that some sections of the Queensland coal industry became concerned that the GC and the operators would no longer be able to cope with a mine fire. They were so attuned to and focussed on looking for tiny concentrations of spontaneous combustion indicator gases that the large concentrations present in a mine fire would render the technique unusable. A subsequent testing program recently established by Simtars showed that these concerns were unfounded and that all Camgas sites were still capable of analysing samples typical of an intense mine fire.